42056 - Human and Object Recognition (ROH) [UB]

Type: S3 Course
Semester: Spring
Teaching Points: 15
Offer: Annual
Responsible Unit: UB
Responsible: Sergio Escalera (UB)
Language: English


This course will review and discuss current approaches to human/object recognition in computer vision. The course will cover bag of words models, classifier based models, concurrent recognition and segmentation, context models for object recognition and human behavior perception. We will be reading a series of papers from computer vision.


See detailed current content here.

1 – Introduction to Human and Object Recognition: the full picture.

2 – Object Recognition
Single object recognition using local features (I): Invariant features (SIFT); Recognition as a classification problem (Ferns);
Single object recognition using local features (II): Alignment (RANSAC); Pose estimation.
From instances to categories: bag of words and visual vocabularies.
Context and scene understanding.
Large Scale Object Recognition: hashing, knowledge transfer.

3 – Human Recognition
Faces: detection (B&W and color), classification (identity, gender, age, etc.).
Facial pose, facial expressions.
Body: upper body detection, full body detection.
Body Pose.
Human action recognition.
Social signals and human behavior.